Mars images rendered in 4K [video]
youtube.com
youtube.com
- The opportunity rover captured a barren scene which was named Rub Al Khali (Yes! On Mars!), named after the Rub Al Khali desert area which touches UAE, Saudi, Oman and Yemen. (1) (2)
- The video says that Mars rovers send back images (not videos) and then those images are stitched together to create videos, because "nothing ever moves on Mars". But I wonder how a Mars video will look. Maybe it will surprise us (movements due to winds).
- Mars has a lot of clay (i.e. it was once a watery planet)
- Some videos took 1,000 images to make
- There is a beautiful selfie at timestamp 6:41
- There is dune named "Namib" on Mars, shown at timestamp 9:10, named after Namib desert in Africa. (3) (4)
Where's the camera in that selfie? Perhaps edited out of the mosiac?
[1] https://www.nasa.gov/feature/jpl/nasas-curiosity-mars-rover-...
That's probably a great example of one of the benefits of these rovers running for so long. They'd never have scheduled something so non-science-oriented if they only got the few weeks they expected, but when your mission exceeds expectations by several thousand percent, you have time for such things.
Here's a dust devil https://www.youtube.com/watch?v=k8lfJ0c7WQ8 (again, it's still images to make a video) but not because "nothing ever moves on Mars"!
Mars is constantly shifting and changing with its weather and seasons. Ice caps freeze more, and ice caps melt more, water or other liquid flows down hills when it gets warm.
More often than dust storms, a new robot comes down to land on Mars in firey rocket-landing-fueled or bouncing-ball style. Movement.
Mars is constantly hit by rocks from space that create craters and push dust up into the atmosphere.
I've never heard someone say "Nothing ever moves on Mars" because it's really not true and a total lack of imagination. Maybe people think nothing moves because we haven't sent video cameras yet? But...Mars is constantly in motion.
I think it has been done? https://www.youtube.com/watch?v=tMntZ6_R78Q
I take a different approach and I can't say I haven't found issues (I have, many, I'm working on them).
My goal is to reach a level where there are almost none artifacts (there are a couple in that video at the start, when the shield is released and when there are many large movements of the camera due to Curiosity dangling off the parachute). As I said, I'm still working on it, and in between my other side projects, work and quarantine I've been on a less-than-ideal situation.
In time, if I pull it off, I'll post it and I guess you'll see it.
> named after Namib desert in Africa
In a few hundred years, maybe these won't be deserts anymore. It will make for an interesting fact!
Crazy to think Earth might eventually look like Mars.
When looking at the title at first I thought they had used an SR-GAN (or modern variant) to upscale the images. And no, they just took the images and pan over them. I wouldn't have liked the SR-GAN approach either, I'm not a fan of it (to me it trades accuracy for good looks, and it introduces some artifacts which, while harder to spot on video, are probably not really there).
Also others have pointed out that it seems like basically monetizing public domain images without putting some effort into it. Like another crows has pointed out, they do work into writing a script, the narrator, and editing the material, so that's not a take I like (and I'm also "in the same business" as them).
The colorization is done by taking several of these images for wavelengths and adding them together. For example blue is in the 450–495 nm wavelength, green 495–570 nm, and red 625–740 nm. The Mastcam on the curiosity rover is actually two cameras with various sensors for different wavelengths (including wavelengths outside the visual range). They can use these images to create color images. You take the 440 nm sensor and give each pixel a value between 0-255 and display that with more/less blue. Take the 525 nm or 550 nm sensor and make them green and the 675 nm sensor and make them red. When you combine these you get an image that looks pretty similar to what it would look if you were really there (It's called true color and false color [1]). This same process is used by satellites as well to measure and classify things on Earth from space. They way different surfaces reflect light in different wavelengths help scientists classify things on Earth and Mars and all over the solar system. (It's Okay to Be Smart on YouTube has a cool video about Infrared [2])
[0] https://mars.nasa.gov/msl/spacecraft/instruments/mastcam/for...
But you are right that many times scientific missions will have monochrome CCDs with a filter wheel on top (Cassini, Galileo, Rosetta, Dawn to name a few).
Calling it “video” may be technically true, but lees interesting than Google Street View (which is not video, but more interactive).
Is this really 4K?
It does look like they're testing it since it's in tech preview [1] so once it's relatively-bug free they'll probably release it.
[1] https://developer.apple.com/safari/technology-preview/releas...
Input #0, matroska,webm, from 'New - Mars In 4K-ZEyAs3NWH4A.f313.webm.part':
Metadata:
encoder : google/video-file
Duration: 00:10:08.41, start: 0.000000, bitrate: 4436 kb/s
Stream #0:0(eng): Video: vp9 (Profile 0), yuv420p(tv, bt709), 3840x2160, SAR 1:1 DAR 16:9, 29.97 fps, 29.97 tbr, 1k tbn, 1k tbc (default)
Can't you use an ad-blocker?
It's basically some pans across high resolution composite images. Quite cool but not captured video. As they mention, there's not much moving there anyway but it will be cool when someday we can look at high def (or even stereoscopic) video from these remote locations.
At one point in the video, there is a 1.8 million pixel image where they discuss and show zooming into the image while preserving detail.
But given how aggressively YouTube compresses videos, watching in 4K can provide a huge improvement over 1080p even if your screen only has, say, 1,600 lines like my MBP does, instead of the 2,160 lines that 4K actually produces. It's really about the clarity of textures and how lifelike they are -- it's very noticeable. Of course, this also assumes you have good vision.
I also don't think anyone's too worried about "wasting computing power" here for a short video.
As far as "wasting computing power"... I hear some people play video games while watching movies... on the same machine. So that's probably what that is about.
Because YouTube (like other streaming services) aggressively compresses both 1080p and 4K, but the 4K version still has a much higher bitrate, watching the 4K stream on a 1080p monitor will be significantly better quality than watching a 1080p stream on a 1080p monitor.
I'm not saying you won't see any compression artifacts at all, of course if it's an scene you will. But the 4K version is still noticeably way better than the 1080p one, even in an action scene. It's a higher bitrate so there's more detail no matter what.
From an information theory perspective, you are obviously correct that there is more information in the 4k streams and therefore you will get more detail from them. My observation however has often been the opposite.
My attempt to rationalize leans towards notions that the lossy compression techniques we use with video codecs still have significant impedance with the "compression" in our neuro-opthalmologic system.
Funnily enough, this reminds me of what I did with JPEGs on websites when Retina first became a thing. Instead of serving up different assets depending on resolution, I discovered that serving up a 2x resolution JPEG to everyone, with really crappy quality, was superior to serving up 1x with high quality. For 1x screens, blockiness was inherently shrunken because of JPEG's fixed block sizes, and for 2x screens the low quality was harder to perceive because it was at a small scale anyways. How compression interacts with scale/resolution is not a necessarily intuitive thing.
Golden rule: don't use Safari for Youtube.
Of course Apple refuses to add VP9; Google demands that Apple would never sue them over any patent. Any. It looks like Google/YouTube caved in, and has/will start encoding (the most recently uploaded) videos in AV1 codec, which will work on tvOS. Hopefully that will translate to Safari on macOS and iOS for those who use it.
For those who had the same question I had about whether these are actual colors, the narrator talks about the change in color between images and the color correction done in post-processing around 5:45.
imagine seeing the swirling dust, shifting sands, the bumps and parallax from a moving rover. that would be so cool!
https://science.nasa.gov/science-news/science-at-nasa/2005/1...
Of course, it's very weak, so it couldn't push over a spacecraft or building like in "The Martian", but I suspect sands shifts on a very slow timescale. I'm not sure.
Apparently the Mars 2020 Rover (https://mars.nasa.gov/mars2020/spacecraft/rover/microphones) will have that capacity when it lands in February 2021 but the microphones may break after landing: "Engineers are optimizing this microphone for space from easily available, store-bought hardware. It is unlikely it will work beyond landing. If it does survive, we may be able to hear the sounds of the Martian winds and sounds of the working rover, such as the wheels turning, or the motors that turn its head, and the heat pumps that keep it warm."
Winds in Mars have almost no strength. Air pressure in the Mars surface is very low, something like 1% of Earths. Dust devils and storms are made of extremely fine powder.
[0] https://youtu.be/k7pfdFMVj-o
[1] They have also breathed new life on one of those (before killing it, yes) by cleaning the solar panels. I'm referring to the MER Rovers Spirit & Opportunity.
The problem with Mars dust is that it's electrostatic and sticks to surfaces like solar panels.
https://www.nasa.gov/feature/goddard/the-fact-and-fiction-of...
>It is unlikely that even these dust storms could strand an astronaut on Mars, however. Even the wind in the largest dust storms likely could not tip or rip apart major mechanical equipment. The winds in the strongest Martian storms top out at about 60 miles per hour, less than half the speed of some hurricane-force winds on Earth.
>Focusing on wind speed may be a little misleading, as well. The atmosphere on Mars is about 1 percent as dense as Earth’s atmosphere. That means to fly a kite on Mars, the wind would need to blow much faster than on Earth to get the kite in the air.
Strongest martian storm can have wind speeds up to 100 km/h. Because low air density, the force of the wind would be equal to "Fresh breeze" in Beaufort Wind Scale. (8.5-10.5 m/s) "Small trees sway, waves break on inland waters."
It's a machine.
Curiosity: https://mars.nasa.gov/msl/multimedia/images (raw: https://mars.nasa.gov/msl/multimedia/raw-images )
If the rover can stream to the orbiter at 2mb/s, would it be possible stream video to the orbiter and have it save in its buffer? Then it'd transfer to Earth later on?
Here's an example on the NASA site: https://www.nasa.gov/image-feature/revealing-what-lies-benea...
"10 minutes" is an ad milestone for videos on youtube. I systematically click away and hide videos that are between 10:00 and 10:20 minutes long: They'll almost certainly have been artificially slowed down to match the ad requirements.
It's infuriating, and once you know about it, you see it everywhere. Anyway, in this instance it matters a lot less as the audio is completely unimportant…
Looking at the channel, 3 of its videos are between 10:04 and 10:09 long. Of course, that's nothing compared to a lot of other long-form youtubers I've seen with like 80% of their videos around that timestamp.
The 10 minutes one seems pretty arbitrary, but to arrive at 8 mins I suspect that Google studies how much time there needs to be so that mid-roll ads don't negatively affect engagement.
(Shout-out to all the wonderful, brilliant engineers working on figuring out how to minimize the user's reaction to cramming AS MANY ADS as possible down their brain)
OK.
I got the point, but it's not really fair towards content creators.